xref: /freebsd/share/man/man9/namei.9 (revision 35c0a8c449fd2b7f75029ebed5e10852240f0865)
1.\"
2.\" Copyright (c) 1998, 1999 Eivind Eklund
3.\" Copyright (c) 2003 Hiten M. Pandya
4.\" Copyright (c) 2005 Robert N. M. Watson
5.\"
6.\" All rights reserved.
7.\"
8.\" This program is free software.
9.\"
10.\" Redistribution and use in source and binary forms, with or without
11.\" modification, are permitted provided that the following conditions
12.\" are met:
13.\" 1. Redistributions of source code must retain the above copyright
14.\"    notice, this list of conditions and the following disclaimer.
15.\" 2. Redistributions in binary form must reproduce the above copyright
16.\"    notice, this list of conditions and the following disclaimer in the
17.\"    documentation and/or other materials provided with the distribution.
18.\"
19.\" THIS SOFTWARE IS PROVIDED BY THE DEVELOPERS ``AS IS'' AND ANY EXPRESS OR
20.\" IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES
21.\" OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE DISCLAIMED.
22.\" IN NO EVENT SHALL THE DEVELOPERS BE LIABLE FOR ANY DIRECT, INDIRECT,
23.\" INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT
24.\" NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
25.\" DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
26.\" THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
27.\" (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF
28.\" THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
29.\"
30.\"
31.\" If you integrate this manpage in another OS, I'd appreciate a note
32.\"	- eivind@FreeBSD.org
33.\"
34.Dd December 17, 2024
35.Dt NAMEI 9
36.Os
37.Sh NAME
38.Nm namei ,
39.Nm NDINIT ,
40.Nm NDFREE
41.Nd pathname translation and lookup operations
42.Sh SYNOPSIS
43.In sys/param.h
44.In sys/fcntl.h
45.In sys/namei.h
46.Ft int
47.Fn namei "struct nameidata *ndp"
48.Ft void
49.Fo NDINIT
50.Fa "struct nameidata *ndp" "u_long op" "u_long flags"
51.Fa "enum uio_seg segflg" "const char *namep" "struct thread *td"
52.Fc
53.Ft void
54.Fn NDFREE "struct nameidata *ndp" "const uint flags"
55.Sh DESCRIPTION
56The
57.Nm
58facility allows the client to perform pathname translation and lookup
59operations.
60The
61.Nm
62functions will increment the reference count for the vnode in question.
63The reference count has to be decremented after use of the vnode, by
64using either
65.Xr vrele 9
66or
67.Xr vput 9 ,
68depending on whether the
69.Dv LOCKLEAF
70flag was specified or not.
71.Pp
72The
73.Fn NDINIT
74function is used to initialize
75.Nm
76components.
77It takes the following arguments:
78.Bl -tag -width ".Fa segflg"
79.It Fa ndp
80The
81.Vt "struct nameidata"
82to initialize.
83.It Fa op
84The operation which
85.Fn namei
86will perform.
87The following operations are valid:
88.Dv LOOKUP , CREATE , DELETE ,
89and
90.Dv RENAME .
91The latter three are just setup for those
92effects; just calling
93.Fn namei
94will not result in
95.Fn VOP_RENAME
96being called.
97.It Fa flags
98Operation flags.
99Several of these can be effective at the same time.
100.It Fa segflg
101UIO segment indicator.
102This indicates if the name of the object is in userspace
103.Pq Dv UIO_USERSPACE
104or in the kernel address space
105.Pq Dv UIO_SYSSPACE .
106.It Fa namep
107Pointer to the component's pathname buffer
108(the file or directory name that will be looked up).
109.It Fa td
110The thread context to use for
111.Nm
112operations and locks.
113.El
114.Sh NAMEI OPERATION FLAGS
115The
116.Fn namei
117function takes the following set of
118.Dq "operation flags"
119that influence its operation:
120.Bl -tag -width ".Dv WANTPARENT"
121.It Dv LOCKLEAF
122Lock vnode on return with
123.Dv LK_EXCLUSIVE
124unless
125.Dv LOCKSHARED
126is also set.
127.Xr VOP_UNLOCK 9
128should be used
129to release the lock (or
130.Xr vput 9
131which is equivalent to calling
132.Xr VOP_UNLOCK 9
133followed by
134.Xr vrele 9 ,
135all in one).
136.It Dv LOCKPARENT
137This flag lets the
138.Fn namei
139function return the parent (directory) vnode,
140.Va ni_dvp
141in locked state, unless it is identical to
142.Va ni_vp ,
143in which case
144.Va ni_dvp
145is not locked per se (but may be locked due to
146.Dv LOCKLEAF ) .
147If a lock is enforced, it should be released using
148.Xr vput 9
149or
150.Xr VOP_UNLOCK 9
151and
152.Xr vrele 9 .
153.It Dv LOCKSHARED
154Lock vnode on return with
155.Dv LK_SHARED ,
156if permitted by the file system that owns the vnode.
157The file system must explicitly permit this by setting
158.Dv MNTK_LOOKUP_SHARED
159in
160.Dv mp->mnt_kern_flag
161during mount and by calling
162.Fn VN_LOCK_ASHARE
163when allocating the vnode.
164If
165.Dv LOCKLEAF
166is specified but shared locking is not permitted, then the vnode will be
167returned with
168.Dv LK_EXCLUSIVE .
169.Xr VOP_UNLOCK 9
170should be used
171to release the lock (or
172.Xr vput 9
173which is equivalent to calling
174.Xr VOP_UNLOCK 9
175followed by
176.Xr vrele 9 ,
177all in one).
178.It Dv WANTPARENT
179This flag allows the
180.Fn namei
181function to return the parent (directory) vnode in an unlocked state.
182The parent vnode must be released separately by using
183.Xr vrele 9 .
184.It Dv NOCACHE
185Avoid
186.Fn namei
187creating this entry in the namecache if it is not
188already present.
189Normally,
190.Fn namei
191will add entries to the name cache
192if they are not already there.
193.It Dv FOLLOW
194With this flag,
195.Fn namei
196will follow the symbolic link if the last part
197of the path supplied is a symbolic link (i.e., it will return a vnode
198for whatever the link points at, instead for the link itself).
199.It Dv NOFOLLOW
200Do not follow symbolic links (pseudo).
201This flag is not looked for by the actual code, which looks for
202.Dv FOLLOW .
203.Dv NOFOLLOW
204is used to indicate to the source code reader that symlinks
205are intentionally not followed.
206.It Dv SAVENAME
207Do not free the pathname buffer at the end of the
208.Fn namei
209invocation; instead, free it later in
210.Fn NDFREE
211so that the caller may access the pathname buffer.
212See below for details.
213.It Dv SAVESTART
214Retain an additional reference to the parent directory; do not free
215the pathname buffer.
216See below for details.
217.El
218.Sh ALLOCATED ELEMENTS
219The
220.Vt nameidata
221structure is composed of the following fields:
222.Bl -tag -width ".Va ni_cnd.cn_pnbuf"
223.It Va ni_startdir
224In the normal case, this is either the current directory or the root.
225It is the current directory if the name passed in does not start with
226.Ql /
227and we have not gone through any symlinks with an absolute path, and
228the root otherwise.
229.Pp
230In this case, it is only used by
231.Fn vfs_lookup ,
232and should not be
233considered valid after a call to
234.Fn namei .
235.It Va ni_dvp
236Vnode pointer to directory of the object on which lookup is performed.
237This is available on successful return if
238.Dv LOCKPARENT
239or
240.Dv WANTPARENT
241is set.
242It is locked if
243.Dv LOCKPARENT
244is set.
245.It Va ni_vp
246Vnode pointer to the resulting object,
247.Dv NULL
248otherwise.
249The
250.Va v_usecount
251field of this vnode is incremented.
252If
253.Dv LOCKLEAF
254is set, it is also locked.
255.Pp
256.It Va ni_cnd.cn_pnbuf
257The pathname buffer contains the location of the file or directory
258that will be used by the
259.Nm
260operations.
261It is managed by the
262.Xr uma 9
263zone allocation interface.
264.El
265.Sh RETURN VALUES
266If successful,
267.Fn namei
268will return 0, otherwise it will return an error.
269.Sh FILES
270.Bl -tag -width Pa
271.It Pa src/sys/kern/vfs_lookup.c
272.El
273.Sh ERRORS
274Errors which
275.Fn namei
276may return:
277.Bl -tag -width Er
278.It Bq Er ENOTDIR
279A component of the specified pathname is not a directory when a directory is
280expected.
281.It Bq Er ENAMETOOLONG
282A component of a pathname exceeded 255 characters,
283or an entire pathname exceeded 1023 characters.
284.It Bq Er ENOENT
285A component of the specified pathname does not exist,
286or the pathname is an empty string.
287.It Bq Er EACCES
288An attempt is made to access a file in a way forbidden by its file access
289permissions.
290.It Bq Er ELOOP
291Too many symbolic links were encountered in translating the pathname.
292.It Bq Er EISDIR
293An attempt is made to open a directory with write mode specified.
294.It Bq Er EINVAL
295The last component of the pathname specified for a
296.Dv DELETE
297or
298.Dv RENAME
299operation is
300.Ql \&. .
301.It Bq Er EROFS
302An attempt is made to modify a file or directory on a read-only file system.
303.El
304.Sh SEE ALSO
305.Xr uio 9 ,
306.Xr uma 9 ,
307.Xr VFS 9 ,
308.Xr vnode 9 ,
309.Xr vput 9 ,
310.Xr vref 9
311.Sh AUTHORS
312.An -nosplit
313This manual page was written by
314.An Eivind Eklund Aq Mt eivind@FreeBSD.org
315and later significantly revised by
316.An Hiten M. Pandya Aq Mt hmp@FreeBSD.org .
317.Sh BUGS
318The
319.Dv LOCKPARENT
320flag does not always result in the parent vnode being locked.
321This results in complications when the
322.Dv LOCKPARENT
323is used.
324In order to solve this for the cases where both
325.Dv LOCKPARENT
326and
327.Dv LOCKLEAF
328are used, it is necessary to resort to recursive locking.
329